@BartekBanachewicz it's hard to remove the shit

@thecoshman If they forced me to use Java, I'd be looking for a new job. Like most jokes, if it has to be explained, it obviously failed. Sorry.

@nwp I can write in whatever really

doesn't mean I won't rant about it -.-

see the problem is that I get I can create working things in python

what really gets to me is the attitude of its community that disagrees with me

when they're obviously wrong

everyone's wrong

even you are wrong

I actually don't mind using Java

I mind using any language really

all of them are bad

all of them

@ratchetfreak I probably should, but I haven't had a reason to touch it in a few years, and don't really expect to very soon either.

@ratchetfreak mmm is ther a mod that could possible move it there? Or should I just copy paste

The main problem is that people are fundamentally bad, so even with a good language, it'd be a pain in the arse to try and work with it

@thecoshman Well, that's the first problem.

@thecoshman but I find some comfort in using languages that are bad just because there weren't enough resources to make them better

If people are going to violate things like SRP so much, they are going to do it not matter what langauge

using languages that are bad because their designers are dumb is just a pita

→ 1 message moved to C++ Questions and Answers

thanks Jerry

like, Lua also has numerous problems and annoyances

but at least I can feel that its creators really tried

and they had a smidge of competence

using python just makes me question human intelligence

@BartekBanachewicz did you know that in print(foo), (Python 2), the trailing comma is here to tell print not to append a newline?

yes

Lua isn't that bad though... a bit funky, but nice overall

sure it isn't

but it's also far from perfect

Aiming for perfection can be a deep rabbit hole

so is Haskell, and Rust, and Terra, and whatever language you can say I "like"

@thecoshman if you're not aiming for perfection you lose no matter what

have you seen shooting competitors aim beside the target because they're gonna miss it anyway

Well... I have compose-able function things now vOv

if your goal never was to make a language with non-retarded semantics, you sure as heck won't make one

@BartekBanachewicz "going to miss anyway"?

@thecoshman the whole "practical" bs

IOW making shortcuts in formal spec to make "print" spell out nicer

and then you get that

what's wrong with lua ?

but yeah, I know what you mean, if you don't even try to make a good language, and just cobble shit together as you need it... you get stuck with shit like PHP and JS

I actually really like Lua, so I guess I'm not the one to ask :P

@thecoshman precisely

it's a fundamentally different approach to design

@Borgleader ooh fancy bullets

And refusal to depreciate means you don't get a new PHP that could have nice unified syntax. And even if you did, refusal to update and move on means you get stuck with python 2 & 3

correct again.

@thecoshman PHP did deprecate a bunch of stuff

this is all because of the shortcuts they make

Lua's multi-arg return is kinda weird tho.

@BartekBanachewicz maybe not making shortcuts is prohibitively expensive

foo[bar()] = 5 # if bar() multi-returned, only the 1st value is used

I think it's great that Google looked at Angular, and basically said "Well... we certainly would do things differently if we started over. Let's start over"

@Ven the "check 2nd param for errors" idiom is kinda silly

@nwp sure, if your goal is to make a language in two weeks

that's how JS was born

Obviously, such antics have to be balanced, you are not going to build a community for a language/library/framework if every release invalidates everything ever. And with out a community it's very had for such things to be used

@nwp only if you are a shortsighted tool

that's why it's super important to flesh out important, hard choices early on

cue Rust early releases

Or make it clear when shit is still wip "so don't write stuff and expect to be able to leave it alone for ever"... cue Rust

I am not saying that language design is easy, but pretending it is and going "I want all the nice syntax" route without tackling the hard problems won't make it so

@thecoshman they kinda did make it clear it's wip though

@Ven my eyes

@Ven actually I'm not sure how my VM would handle that

check and fix it

what do you mean by "fix"?

it's probably broken in your VM :P

maybe because your VM is broken

@BartekBanachewicz my point

@Ven in your face, bitch

:(

over focus on "nice things" tends to leave a lot of warts in things you aren't thinking about in the design fase

@Ven although I will give you that I should really newtype the result list to get rid of that ugly partial head

(:[])

what :D

inb4 "OMG OPERATORS"

from me?

cmon

nah, surely not from you

you made me realize that I actually miss working on Turnip

that's not my school :P

I just wish it wasn't so useless

well

how much of Lua does it run

@Ven is that a "design fase" or an "ugly partial head"

@Ven about that much

@nwp it's a pure

@BartekBanachewicz just didn't want to write pure or return

ah, face-lift, right

@Ven I'm not particularly fond of monad/app instance for []

call me prejudiced

well, pure/returnfor [] is easy enough to understand. (:[]) is a bit funny looking ;-)

eh, for one if I actually did that newtype, that pure could stay there

but that's about it really

no I mean

(:[]) is strictly equal to pure

as long as we're talking []. What I had in mind was a newtype/data to get rid of head, which, if made an instance of App, would still accept pure to construct it

so my solution is actually better on all points

: ^)

yep.

ven-so-functional

nevermind that I've only written Haskell twice in my life

I'm the best at hoogle

BinOP call

it should dispatch to a metatable call, right?

oh, yeah. I just haven't implemented metatables yet.

@Ven operators are kind of a hack atm

wtf dude

so many monaderrors

that's ugly as fuck :c you could technically have an AST with an operator called for 3 operands >.>

@Ven Well if Lua could do e.g. (+)(a,b,c), yeah. But that won't parse.

;-)

can you change an integer's metatable?

I can never remember that actually

But I don't think Lua is particularly clever in its implementation either.

@Ven also, well, not really. AST.BinOp name lhs rhs kinda doesn't allow that. It's just when it's passed through eval.

your types allow that. bad! :)

it's just because I resort to eval (Call ...) to call the operator

myeah

not gonna bore you on how I'd do that. Or on how I implemented my lisp in perl ;P

since eval doesn't do the same in both, by far. It should. Terra probably does

@Ven See, so technically my implementation isn't bad per se, it's just not really super close to Lua operator semantics :P

yeah..

the nitty-gritty are always dangerous.(or is?)